抄録
One aspect of formulating chunks in EFL silent reading was examined, controlling visual inputs. Three written texts were displayed with Japanese university students leaning English as FL (N=27) to determine whether the presentation modes (i.e. clause-, phrase- and wordmodes) affect their comprehension rate and overall processing time. It was found that, while neither clause-by-clause nor phrase-by-phrase presentation interfered with comprehension, word-by-word mode of presentation served as a negative factor for the construction of processing units in coding written input. The results seemto reveal that Japanese EFL readers, as have been suggested by Kadota (1982) and Kadota and Tada (1992), process printed sentences in terms of phrase- or clause-like chunks rather than words.
